Geoscience Reference
In-Depth Information
and grass 11.2%. Forest and impervious were excluded from the model by SPSS
due to colinearity and insignificance of the variables. Although the model shows
that the four variables negatively influence IBI scores, the model was skewed by
outlying (
0.01%) values of wet vegetation. When the water/wet vegetation class
was excluded from the model only agriculture was a significant predictor of IBI
score (adjusted
R
2
<
0.162). The results of regression using the dependent variable
QHEI show a negative influence of agriculture on QHEI score, with agriculture
explaining 22.3% (adjusted
R
2
) of the variance in QHEI scores. The remaining land
cover variables were also excluded from the habitat model due to colinearity and
insignificance of the variables. No significant relationships were found for either IBI
score or QHEI score within a 100 m buffer of the sample points. Additionally, the
bivariate regression analysis of IBI and QHEI scores revealed no relationship between
biotic integrity and habitat scores using the Tribs subset of the larger OEPA dataset.
ΒΌ
10.5 Discussion
Although multiple regression analysis was largely inconclusive, agricultural land
uses do appear to have greatly impacted the health of the Tribs. Larger sample sizes
would certainly have yielded more reliable results. Despite the inability of theses
analyses to point with certainty toward a particular land use culprit, examination of
the summary statistics against land cover proportions gives insight into possible
trends. In an area with predominantly agricultural land use combined with extre-
mely low biotic integrity, even a loose correlation gives cause for examination of
the practices that have likely contributed to declining fish communities.
Overall, habitat scores relative to the other streams in the Maumee AOC were
even worse than IBI scores. The Tribs' low QHEI scores can be attributable to
extensive channelization, complete removal of riparian vegetation, tile drainage
and excessive sedimentation, all of which can be traced back to extensive agricultural
practices within the watersheds. Agricultural fields, at various locations throughout
the Tribs watersheds, are cultivated to within less than a meter of streams and
drainage ditches. The lack of vegetation in these areas means lost habitat for macro-
invertebrates and fish. Furthermore, these practices input large amounts of sediments
to streams. The Ohio EPA assessments determined that siltation is heavy to moderate
and silt is the dominant substrate type for all sites in the study area. Most sites also
exhibit recent or sustained channelization with a few sites in stages of recovery from
channelization. In contrast, many of the urban and residential sites in the Maumee
AOC have no channelization (for example lower reaches of Tenmile Creek/ Ottawa
River and Grassy Creek).
Comparison of IBI summary statistics between the study area and the entire
Maumee AOC indicate that nearly homogeneous agricultural land cover may be
just as detrimental as or worse for stream biotic integrity than urban land uses.
In addition to the habitat impacts mentioned above, agricultural pollutants such as
nutrients, pesticides and dissolved organic matter are also a likely cause of
